Project 1A: To investigate mucosa-associated alterations of the lung microbiome during carcinogenesis, we utilized the NCI-MD case-control study and sequenced the V3-V5 16S rRNA gene. Candidate bacterial identities were validated in lung cancer samples from The Cancer Genome Atlas. We examined the ecological diversity within samples (alpha diversity) and between samples (beta diversity) of normal healthy tissues and NCI-MD non-tumor adjacent and tumor tissues. We found increasing diversity and richness in subjects with lung cancer, which could be related to changes in the lung microenvironment, indicating acquisition of somatic mutations or alterations in the immune system. We identified two genera that were able to classify lung cancer cases from immediate autopsy controls, Variovorax and Streptococcus, and may be biomarkers of lung cancer risk or associated with cancer initiation and progression. Variovorax abundance was higher in tumor compared to non-tumor tissues from the same individual, a finding that was validated in the larger TCGA dataset. Using the Resphera Insight to speciate our 16S rRNA gene reads, we were able to putatively identify the species as V. paradoxus. Fluorescent in situ hybridization (FISH) analysis of resected lung tumors further confirmed the presence of V. paradoxus in tumor tissue. V.paradoxus was localized intracellularly in tumor cells. A higher abundance of Variovorax and lower abundance of Streptococcus was seen in Stage I lung cancer, suggesting these maybe biomarkers of cancer risk or associated with cancer initiation. Variovorax species could differentiate between histological subtypes of lung cancer, adenocarcinoma (AD) and squamous cell carcinoma (SCC). We hypothesized that SCC-associated microbial taxa would be more abundant in tumors with TP53 mutations.. We investigated the association between TP53 mutations, microbial taxa, and tumor histology. Specific microbial taxa were more abundant in SCC with TP53 mutations, whereas no pattern of association was found among AD with TP53 mutations. These data support the hypothesis that in tumors with TP53 mutations, SCC-associated taxa gain greater access to the tumor microenvironment.. Species-level validation and quantification will be achieved by droplet digital PCR (ddPCR) methods. We are developing methods to identify bacterial isolates from lung tumors and are examining the role of the mucosa-associated lung microbiome, its relationship to inflammation and the metabolome in lung cancer, and the mechanisms underlying this relationship. We will explore these mechanistic relationship between specific bacterial species using a K-ras/p53R172H mouse model in the NCI gnotobiotic mouse facility and in the antibiotic-treated SPF mice. Project 1B: We developed methylation-specific ddPCR to detect and quantify rare methylation events to investigate biomarkers in liquid biopsies. The ddPCR assay could detect as few as 30 haploid genome-equivalents of methylated promoter DNA, and count a single methylated allele present at 0.2%. Differences in methylation levels between tumors and adjacent tissues were also observed. Therefore, we have established a robust and ultrasensitive method for standardized determination of DNA promoter methylation status. We are now evaluating the potential value of the ddPCR assay for prognostic classification of lung cancer patients using FFPE tumor samples. We obtained preliminary evidence that methylated DNA is detectable in ctDNA from lung cancer patients at advanced stage, and are continuing to optimize conditions for detection in patients with Stage I lung cancer. Project 1C: Another project was to evaluate urinary and blood tumor metabolite liquid biopsy approaches for early detection of lung and liver cancer and for assessing patient outcome. In a collaboration supported by the Department of Defense (DoD) Clinical Exploration Award, we are in the process of investigating the predictive capability of urinarymetabolites in the Detection of Early lung Cancer among Military Personnel (DECAMP) consortium and the Walter Reed Naval Medical Center (WRNMMC) LDCT Scan Lung Cancer Screening Program. In this ongoing acruement, we have received cases and controls consisting of former or current smokers with an indeterminate pulmonary nodule 7-30 mm in size on chest CT within the year prior to enrollment. They were then followed with standard of care for up to 2 years until a final diagnosis of lung cancer or benign lung disease was made. Colleagues have shown that gene expression features measured in bronchial epithelial cells and the more distant nasal epithelium of smokers reflect the presence of cancer in the lung, consistent with the presence of a molecular field of injury caused by cigarette smoke. Our metabolomic analysis will be compared to, and integrated with, the analysis of bronchial and nasal epithelial cells from these patients. We have received urine, blood and tissue samples from WRNMMC. We will confirm the association of 4-urinary metabolites with lung cancer diagnosis and prognosis, investigate what urinary and blood biomarkers result from altered tumor tissue metabolites, and whether they are associated with histological subgroups, or driven by a specific mutation. We have received lung cancer cases and controls from the LDCT arm of the NLST and are in the process of measuring 4 metabolites in the urine of these subjects to answer the aforementioned questions. We are in the process of evaluating whether urinary tumor metabolites CR and NANA decrease after surgery in the Lung Cancer Biospecimen Resource Network (LCBRN) samples. We have received pre-, 6-, 12- and 24-months post-surgery urine specimens from 46 patients, with follow-up recurrence and survival data. We also obtained 20-month post-surgery urine specimens from 11 patients, and 24-month from 21 patients. We plan to evaluate whether elevated urinary levels of these metabolites are associated with lung cancer recurrence. Urinary metabolites are associated with liver cancer diagnosis and validated in the NCI-UMD and TIGER LC cohorts. We are investigating if the 4 urinary metabolites discovered in our lung cancer studies are also diagnositic of liver cancer. We initially conducted a pilot study in 95 hepatocellular carcinoma (HCC) cases, in collaboration with Xin Wang, LHC and Chulabhorn Research Institute. Controls and high-risk subjects were frequency matched to cases on age, gender and race. This study showed that urinary tumor metabolites previously identified as increased in lung cancer are also elevated in HCC cases compared to population controls and, more importantly, high risk subjects. Next, we evaluated whether the same results were observed in the independent Thailand Initiative in Genomics and Expression Research for Liver Cancer (TIGER-LC) cohort. We hypothesized that the 4 metabolites are also elevated in cholangiocarcinoma (CCA) in addition to HCC. The results of this analysis indicated that urinary tumor metabolites are significantly elevated in CCA and HCC when compared to their respective high risk subjects and unaffected controls.Multivariable analysis adjusted for age, gender, HBV and HCV status indicates that the predictive ability of 4 metabolites to diagnose HCC was high when compared to high-risk subjects. Similar results were observed in CCA. Additionally, levels of 3 metabolites (creatine riboside, N-acetyulneuraminic acid and metabolite 561+), were significantly higher in CCA when compared to HCC. Urinary 4-metabolite profile shows a high predictive ability to diagnose CCA when compared to a clinically utilize tumor marker CA19-9, while their combination leads to a significantly improved classifier pointing to a clinical utility of this type of liquid biopsy.